@llamaindex/liteparse 2.3.0 → 2.4.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/lib.d.ts +25 -0
- package/dist/lib.d.ts.map +1 -1
- package/dist/lib.js +5 -0
- package/dist/lib.js.map +1 -1
- package/dist/native.d.ts +10 -0
- package/dist/native.d.ts.map +1 -1
- package/dist/native.js.map +1 -1
- package/liteparse.linux-x64-gnu.node +0 -0
- package/package.json +8 -8
package/dist/lib.d.ts
CHANGED
|
@@ -34,6 +34,23 @@ export interface LiteParseConfig {
|
|
|
34
34
|
* OCR pod, at the cost of extra OCR-server load. HTTP OCR engine only.
|
|
35
35
|
*/
|
|
36
36
|
ocrHedgeDelaysMs: number[];
|
|
37
|
+
/**
|
|
38
|
+
* Emit per-word sub-boxes on each text item ({@link TextItem.words}).
|
|
39
|
+
* Default false. Word boxes roughly double the text-item payload (size + napi
|
|
40
|
+
* marshalling), so enable only when doing word-level bbox attribution.
|
|
41
|
+
*/
|
|
42
|
+
emitWordBoxes: boolean;
|
|
43
|
+
}
|
|
44
|
+
/**
|
|
45
|
+
* One word's bounding box within a {@link TextItem}, in the same viewport space
|
|
46
|
+
* (top-left origin, 72 DPI). `text` excludes inter-word spaces.
|
|
47
|
+
*/
|
|
48
|
+
export interface WordBox {
|
|
49
|
+
text: string;
|
|
50
|
+
x: number;
|
|
51
|
+
y: number;
|
|
52
|
+
width: number;
|
|
53
|
+
height: number;
|
|
37
54
|
}
|
|
38
55
|
export interface TextItem {
|
|
39
56
|
text: string;
|
|
@@ -46,6 +63,13 @@ export interface TextItem {
|
|
|
46
63
|
confidence?: number;
|
|
47
64
|
/** Rotation in degrees (viewport space). Defaults to 0 when omitted. */
|
|
48
65
|
rotation?: number;
|
|
66
|
+
/**
|
|
67
|
+
* Per-word sub-boxes within this item. A text item groups several words
|
|
68
|
+
* together (breaking only at line/column boundaries), so this carries the
|
|
69
|
+
* finer word-level geometry for bbox attribution. Empty/undefined for items
|
|
70
|
+
* with no word split (e.g. OCR-sourced or single-token items).
|
|
71
|
+
*/
|
|
72
|
+
words?: WordBox[];
|
|
49
73
|
}
|
|
50
74
|
/**
|
|
51
75
|
* A vector-graphic primitive supplied to {@link LiteParse.parsePages}. `kind`
|
|
@@ -88,6 +112,7 @@ export interface ParsedPage {
|
|
|
88
112
|
width: number;
|
|
89
113
|
height: number;
|
|
90
114
|
text: string;
|
|
115
|
+
markdown: string;
|
|
91
116
|
textItems: TextItem[];
|
|
92
117
|
}
|
|
93
118
|
export interface ExtractedImage {
|
package/dist/lib.d.ts.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"lib.d.ts","sourceRoot":"","sources":["../src/lib.ts"],"names":[],"mappings":"AAgBA,MAAM,MAAM,cAAc,GAAG,MAAM,GAAG,MAAM,GAAG,UAAU,CAAC;AAC1D,MAAM,MAAM,YAAY,GAAG,MAAM,GAAG,MAAM,GAAG,UAAU,CAAC;AACxD,MAAM,MAAM,SAAS,GAAG,KAAK,GAAG,aAAa,GAAG,OAAO,CAAC;AAExD,MAAM,WAAW,eAAe;IAC9B,WAAW,EAAE,MAAM,CAAC;IACpB,UAAU,EAAE,OAAO,CAAC;IACpB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,oEAAoE;IACpE,gBAAgB,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAC1C,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,QAAQ,EAAE,MAAM,CAAC;IACjB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,GAAG,EAAE,MAAM,CAAC;IACZ,YAAY,EAAE,YAAY,CAAC;IAC3B,gFAAgF;IAChF,SAAS,EAAE,SAAS,CAAC;IACrB,wFAAwF;IACxF,YAAY,EAAE,OAAO,CAAC;IACtB,qBAAqB,EAAE,OAAO,CAAC;IAC/B,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,KAAK,EAAE,OAAO,CAAC;IACf,UAAU,EAAE,MAAM,CAAC;IACnB;;;;;OAKG;IACH,eAAe,EAAE,OAAO,CAAC;IACzB;;;;;OAKG;IACH,gBAAgB,EAAE,MAAM,EAAE,CAAC;
|
|
1
|
+
{"version":3,"file":"lib.d.ts","sourceRoot":"","sources":["../src/lib.ts"],"names":[],"mappings":"AAgBA,MAAM,MAAM,cAAc,GAAG,MAAM,GAAG,MAAM,GAAG,UAAU,CAAC;AAC1D,MAAM,MAAM,YAAY,GAAG,MAAM,GAAG,MAAM,GAAG,UAAU,CAAC;AACxD,MAAM,MAAM,SAAS,GAAG,KAAK,GAAG,aAAa,GAAG,OAAO,CAAC;AAExD,MAAM,WAAW,eAAe;IAC9B,WAAW,EAAE,MAAM,CAAC;IACpB,UAAU,EAAE,OAAO,CAAC;IACpB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,oEAAoE;IACpE,gBAAgB,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAC1C,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,QAAQ,EAAE,MAAM,CAAC;IACjB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,GAAG,EAAE,MAAM,CAAC;IACZ,YAAY,EAAE,YAAY,CAAC;IAC3B,gFAAgF;IAChF,SAAS,EAAE,SAAS,CAAC;IACrB,wFAAwF;IACxF,YAAY,EAAE,OAAO,CAAC;IACtB,qBAAqB,EAAE,OAAO,CAAC;IAC/B,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,KAAK,EAAE,OAAO,CAAC;IACf,UAAU,EAAE,MAAM,CAAC;IACnB;;;;;OAKG;IACH,eAAe,EAAE,OAAO,CAAC;IACzB;;;;;OAKG;IACH,gBAAgB,EAAE,MAAM,EAAE,CAAC;IAC3B;;;;OAIG;IACH,aAAa,EAAE,OAAO,CAAC;CACxB;AAED;;;GAGG;AACH,MAAM,WAAW,OAAO;IACtB,IAAI,EAAE,MAAM,CAAC;IACb,CAAC,EAAE,MAAM,CAAC;IACV,CAAC,EAAE,MAAM,CAAC;IACV,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,MAAM,CAAC;CAChB;AAED,MAAM,WAAW,QAAQ;IACvB,IAAI,EAAE,MAAM,CAAC;IACb,CAAC,EAAE,MAAM,CAAC;IACV,CAAC,EAAE,MAAM,CAAC;IACV,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,MAAM,CAAC;IACf,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,wEAAwE;IACxE,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB;;;;;OAKG;IACH,KAAK,CAAC,EAAE,OAAO,EAAE,CAAC;CACnB;AAED;;;;;;;GAOG;AACH,MAAM,WAAW,OAAO;IACtB,IAAI,EAAE,QAAQ,GAAG,MAAM,CAAC;IACxB,EAAE,CAAC,EAAE,MAAM,CAAC;IACZ,EAAE,CAAC,EAAE,MAAM,CAAC;IACZ,EAAE,CAAC,EAAE,MAAM,CAAC;IACZ,EAAE,CAAC,EAAE,MAAM,CAAC;IACZ,CAAC,CAAC,EAAE,MAAM,CAAC;IACX,CAAC,CAAC,EAAE,MAAM,CAAC;IACX,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,SAAS,CAAC,EAAE,OAAO,CAAC;IACpB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED;;;;GAIG;AACH,MAAM,WAAW,SAAS;IACxB,UAAU,EAAE,MAAM,CAAC;IACnB,SAAS,EAAE,MAAM,CAAC;IAClB,UAAU,EAAE,MAAM,CAAC;IACnB,SAAS,EAAE,QAAQ,EAAE,CAAC;IACtB,QAAQ,CAAC,EAAE,OAAO,EAAE,CAAC;CACtB;AAED,MAAM,WAAW,UAAU;IACzB,OAAO,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,MAAM,CAAC;IACf,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,EAAE,MAAM,CAAC;IACjB,SAAS,EAAE,QAAQ,EAAE,CAAC;CACvB;AAED,MAAM,WAAW,cAAc;IAC7B,wFAAwF;IACxF,EAAE,EAAE,MAAM,CAAC;IACX,IAAI,EAAE,MAAM,CAAC;IACb,MAAM,EAAE,MAAM,CAAC;IACf,KAAK,EAAE,MAAM,CAAC;CACf;AAED,MAAM,WAAW,WAAW;IAC1B,KAAK,EAAE,UAAU,EAAE,CAAC;IACpB,IAAI,EAAE,MAAM,CAAC;IACb,gEAAgE;IAChE,MAAM,EAAE,cAAc,EAAE,CAAC;CAC1B;AAED,MAAM,WAAW,gBAAgB;IAC/B,OAAO,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,MAAM,CAAC;IACf,WAAW,EAAE,MAAM,CAAC;CACrB;AAED;;;GAGG;AACH,MAAM,WAAW,mBAAmB;IAClC,UAAU,EAAE,MAAM,CAAC;IACnB,UAAU,EAAE,MAAM,CAAC;IACnB,8DAA8D;IAC9D,YAAY,EAAE,MAAM,CAAC;IACrB,oBAAoB,EAAE,OAAO,CAAC;IAC9B,eAAe,EAAE,MAAM,CAAC;IACxB,2DAA2D;IAC3D,aAAa,EAAE,MAAM,CAAC;IACtB,0EAA0E;IAC1E,oBAAoB,EAAE,MAAM,CAAC;IAC7B;;;;OAIG;IACH,aAAa,EAAE,OAAO,CAAC;IACvB;;;OAGG;IACH,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAC7B,SAAS,EAAE,OAAO,CAAC;IACnB,QAAQ,EAAE,MAAM,CAAC;IACjB,2EAA2E;IAC3E,QAAQ,EAAE,OAAO,CAAC;IAClB;;;;OAIG;IACH,OAAO,EAAE,MAAM,EAAE,CAAC;CACnB;AAMD,qBAAa,SAAS;IACpB,OAAO,CAAC,OAAO,CAAkB;IACjC,OAAO,CAAC,OAAO,CAAkB;gBAErB,UAAU,GAAE,OAAO,CAAC,eAAe,CAAM;IAgD/C,KAAK,CAAC,KAAK,EAAE,cAAc,GAAG,OAAO,CAAC,WAAW,CAAC;IAYxD;;;;;OAKG;IACH,UAAU,CAAC,KAAK,EAAE,SAAS,EAAE,GAAG,WAAW;IAgB3C;;;;OAIG;IACG,SAAS,CAAC,KAAK,EAAE,cAAc,GAAG,OAAO,CAAC,mBAAmB,EAAE,CAAC;IAsBhE,UAAU,CACd,KAAK,EAAE,cAAc,EACrB,WAAW,CAAC,EAAE,MAAM,EAAE,GACrB,OAAO,CAAC,gBAAgB,EAAE,CAAC;IAe9B,SAAS,IAAI,eAAe;CAG7B;AAyCD,MAAM,WAAW,kBAAkB;IACjC,MAAM,EAAE,MAAM,CAAC;IACf,aAAa,CAAC,EAAE,OAAO,CAAC;CACzB;AAED,wBAAgB,WAAW,CACzB,KAAK,EAAE,QAAQ,EAAE,EACjB,OAAO,EAAE,kBAAkB,GAC1B,QAAQ,EAAE,CAOZ;AAED,eAAe,SAAS,CAAC"}
|
package/dist/lib.js
CHANGED
|
@@ -24,6 +24,7 @@ export class LiteParse {
|
|
|
24
24
|
numWorkers: userConfig.numWorkers,
|
|
25
25
|
ocrFailureFatal: userConfig.ocrFailureFatal,
|
|
26
26
|
ocrHedgeDelaysMs: userConfig.ocrHedgeDelaysMs,
|
|
27
|
+
emitWordBoxes: userConfig.emitWordBoxes,
|
|
27
28
|
};
|
|
28
29
|
this._native = new native.LiteParse(nativeConfig);
|
|
29
30
|
// Read back the resolved config from the native side
|
|
@@ -46,6 +47,7 @@ export class LiteParse {
|
|
|
46
47
|
numWorkers: resolved.numWorkers ?? 1,
|
|
47
48
|
ocrFailureFatal: resolved.ocrFailureFatal ?? true,
|
|
48
49
|
ocrHedgeDelaysMs: resolved.ocrHedgeDelaysMs ?? [],
|
|
50
|
+
emitWordBoxes: resolved.emitWordBoxes ?? false,
|
|
49
51
|
};
|
|
50
52
|
}
|
|
51
53
|
async parse(input) {
|
|
@@ -123,6 +125,7 @@ function toPage(p) {
|
|
|
123
125
|
width: p.width,
|
|
124
126
|
height: p.height,
|
|
125
127
|
text: p.text,
|
|
128
|
+
markdown: p.markdown,
|
|
126
129
|
textItems: p.textItems.map(toTextItem),
|
|
127
130
|
};
|
|
128
131
|
}
|
|
@@ -144,6 +147,8 @@ function toTextItem(item) {
|
|
|
144
147
|
fontName: item.fontName,
|
|
145
148
|
fontSize: item.fontSize,
|
|
146
149
|
confidence: item.confidence,
|
|
150
|
+
rotation: item.rotation,
|
|
151
|
+
words: item.words,
|
|
147
152
|
};
|
|
148
153
|
}
|
|
149
154
|
export function searchItems(items, options) {
|
package/dist/lib.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"lib.js","sourceRoot":"","sources":["../src/lib.ts"],"names":[],"mappings":"AAAA,OAAO,EACL,MAAM,GASP,MAAM,aAAa,CAAC;
|
|
1
|
+
{"version":3,"file":"lib.js","sourceRoot":"","sources":["../src/lib.ts"],"names":[],"mappings":"AAAA,OAAO,EACL,MAAM,GASP,MAAM,aAAa,CAAC;AA8LrB,8EAA8E;AAC9E,kBAAkB;AAClB,8EAA8E;AAE9E,MAAM,OAAO,SAAS;IACZ,OAAO,CAAkB;IACzB,OAAO,CAAkB;IAEjC,YAAY,aAAuC,EAAE;QACnD,MAAM,YAAY,GAA0B;YAC1C,WAAW,EAAE,UAAU,CAAC,WAAW;YACnC,UAAU,EAAE,UAAU,CAAC,UAAU;YACjC,YAAY,EAAE,UAAU,CAAC,YAAY;YACrC,gBAAgB,EAAE,UAAU,CAAC,gBAAgB;YAC7C,YAAY,EAAE,UAAU,CAAC,YAAY;YACrC,QAAQ,EAAE,UAAU,CAAC,QAAQ;YAC7B,WAAW,EAAE,UAAU,CAAC,WAAW;YACnC,GAAG,EAAE,UAAU,CAAC,GAAG;YACnB,YAAY,EAAE,UAAU,CAAC,YAAY;YACrC,SAAS,EAAE,UAAU,CAAC,SAAS;YAC/B,YAAY,EAAE,UAAU,CAAC,YAAY;YACrC,qBAAqB,EAAE,UAAU,CAAC,qBAAqB;YACvD,QAAQ,EAAE,UAAU,CAAC,QAAQ;YAC7B,KAAK,EAAE,UAAU,CAAC,KAAK;YACvB,UAAU,EAAE,UAAU,CAAC,UAAU;YACjC,eAAe,EAAE,UAAU,CAAC,eAAe;YAC3C,gBAAgB,EAAE,UAAU,CAAC,gBAAgB;YAC7C,aAAa,EAAE,UAAU,CAAC,aAAa;SACxC,CAAC;QAEF,IAAI,CAAC,OAAO,GAAG,IAAI,MAAM,CAAC,SAAS,CAAC,YAAY,CAAC,CAAC;QAElD,qDAAqD;QACrD,MAAM,QAAQ,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC;QACrC,IAAI,CAAC,OAAO,GAAG;YACb,WAAW,EAAE,QAAQ,CAAC,WAAW,IAAI,KAAK;YAC1C,UAAU,EAAE,QAAQ,CAAC,UAAU,IAAI,IAAI;YACvC,YAAY,EAAE,QAAQ,CAAC,YAAY,IAAI,SAAS;YAChD,gBAAgB,EAAE,QAAQ,CAAC,gBAAgB,IAAI,SAAS;YACxD,YAAY,EAAE,QAAQ,CAAC,YAAY,IAAI,SAAS;YAChD,QAAQ,EAAE,QAAQ,CAAC,QAAQ,IAAI,IAAI;YACnC,WAAW,EAAE,QAAQ,CAAC,WAAW,IAAI,SAAS;YAC9C,GAAG,EAAE,QAAQ,CAAC,GAAG,IAAI,GAAG;YACxB,YAAY,EAAG,QAAQ,CAAC,YAA6B,IAAI,MAAM;YAC/D,SAAS,EAAG,QAAQ,CAAC,SAAuB,IAAI,aAAa;YAC7D,YAAY,EAAE,QAAQ,CAAC,YAAY,IAAI,IAAI;YAC3C,qBAAqB,EAAE,QAAQ,CAAC,qBAAqB,IAAI,KAAK;YAC9D,QAAQ,EAAE,QAAQ,CAAC,QAAQ,IAAI,SAAS;YACxC,KAAK,EAAE,QAAQ,CAAC,KAAK,IAAI,KAAK;YAC9B,UAAU,EAAE,QAAQ,CAAC,UAAU,IAAI,CAAC;YACpC,eAAe,EAAE,QAAQ,CAAC,eAAe,IAAI,IAAI;YACjD,gBAAgB,EAAE,QAAQ,CAAC,gBAAgB,IAAI,EAAE;YACjD,aAAa,EAAE,QAAQ,CAAC,aAAa,IAAI,KAAK;SAC/C,CAAC;IACJ,CAAC;IAED,KAAK,CAAC,KAAK,CAAC,KAAqB;QAC/B,mDAAmD;QACnD,MAAM,WAAW,GACf,OAAO,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACzD,MAAM,MAAM,GAAsB,MAAM,IAAI,CAAC,OAAO,CAAC,KAAK,CAAC,WAAW,CAAC,CAAC;QACxE,OAAO;YACL,KAAK,EAAE,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,MAAM,CAAC;YAC/B,IAAI,EAAE,MAAM,CAAC,IAAI;YACjB,MAAM,EAAE,CAAC,MAAM,CAAC,MAAM,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC;SAC3C,CAAC;IACJ,CAAC;IAED;;;;;OAKG;IACH,UAAU,CAAC,KAAkB;QAC3B,MAAM,WAAW,GAAsB,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;YACvD,UAAU,EAAE,CAAC,CAAC,UAAU;YACxB,SAAS,EAAE,CAAC,CAAC,SAAS;YACtB,UAAU,EAAE,CAAC,CAAC,UAAU;YACxB,SAAS,EAAE,CAAC,CAAC,SAAS;YACtB,QAAQ,EAAE,CAAC,CAAC,QAAQ;SACrB,CAAC,CAAC,CAAC;QACJ,MAAM,MAAM,GAAG,IAAI,CAAC,OAAO,CAAC,UAAU,CAAC,WAAW,CAAC,CAAC;QACpD,OAAO;YACL,KAAK,EAAE,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,MAAM,CAAC;YAC/B,IAAI,EAAE,MAAM,CAAC,IAAI;YACjB,MAAM,EAAE,CAAC,MAAM,CAAC,MAAM,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC;SAC3C,CAAC;IACJ,CAAC;IAED;;;;OAIG;IACH,KAAK,CAAC,SAAS,CAAC,KAAqB;QACnC,MAAM,WAAW,GACf,OAAO,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACzD,MAAM,KAAK,GACT,MAAM,IAAI,CAAC,OAAO,CAAC,SAAS,CAAC,WAAW,CAAC,CAAC;QAC5C,OAAO,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;YACvB,UAAU,EAAE,CAAC,CAAC,UAAU;YACxB,UAAU,EAAE,CAAC,CAAC,UAAU;YACxB,YAAY,EAAE,CAAC,CAAC,YAAY;YAC5B,oBAAoB,EAAE,CAAC,CAAC,oBAAoB;YAC5C,eAAe,EAAE,CAAC,CAAC,eAAe;YAClC,aAAa,EAAE,CAAC,CAAC,aAAa;YAC9B,oBAAoB,EAAE,CAAC,CAAC,oBAAoB;YAC5C,aAAa,EAAE,CAAC,CAAC,aAAa;YAC9B,mBAAmB,EAAE,CAAC,CAAC,mBAAmB,IAAI,SAAS;YACvD,SAAS,EAAE,CAAC,CAAC,SAAS;YACtB,QAAQ,EAAE,CAAC,CAAC,QAAQ;YACpB,QAAQ,EAAE,CAAC,CAAC,QAAQ;YACpB,OAAO,EAAE,CAAC,CAAC,OAAO;SACnB,CAAC,CAAC,CAAC;IACN,CAAC;IAED,KAAK,CAAC,UAAU,CACd,KAAqB,EACrB,WAAsB;QAEtB,MAAM,WAAW,GACf,OAAO,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACzD,MAAM,OAAO,GAAG,MAAM,IAAI,CAAC,OAAO,CAAC,UAAU,CAC3C,WAAW,EACX,WAAW,IAAI,IAAI,CACpB,CAAC;QACF,OAAO,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;YACzB,OAAO,EAAE,CAAC,CAAC,OAAO;YAClB,KAAK,EAAE,CAAC,CAAC,KAAK;YACd,MAAM,EAAE,CAAC,CAAC,MAAM;YAChB,WAAW,EAAE,CAAC,CAAC,WAAW;SAC3B,CAAC,CAAC,CAAC;IACN,CAAC;IAED,SAAS;QACP,OAAO,EAAE,GAAG,IAAI,CAAC,OAAO,EAAE,CAAC;IAC7B,CAAC;CACF;AAED,SAAS,MAAM,CAAC,CAAmB;IACjC,OAAO;QACL,OAAO,EAAE,CAAC,CAAC,OAAO;QAClB,KAAK,EAAE,CAAC,CAAC,KAAK;QACd,MAAM,EAAE,CAAC,CAAC,MAAM;QAChB,IAAI,EAAE,CAAC,CAAC,IAAI;QACZ,QAAQ,EAAE,CAAC,CAAC,QAAQ;QACpB,SAAS,EAAE,CAAC,CAAC,SAAS,CAAC,GAAG,CAAC,UAAU,CAAC;KACvC,CAAC;AACJ,CAAC;AAED,SAAS,OAAO,CAAC,GAAyB;IACxC,OAAO;QACL,EAAE,EAAE,GAAG,CAAC,EAAE;QACV,IAAI,EAAE,GAAG,CAAC,IAAI;QACd,MAAM,EAAE,GAAG,CAAC,MAAM;QAClB,KAAK,EAAE,GAAG,CAAC,KAAK;KACjB,CAAC;AACJ,CAAC;AAED,SAAS,UAAU,CAAC,IAAoB;IACtC,OAAO;QACL,IAAI,EAAE,IAAI,CAAC,IAAI;QACf,CAAC,EAAE,IAAI,CAAC,CAAC;QACT,CAAC,EAAE,IAAI,CAAC,CAAC;QACT,KAAK,EAAE,IAAI,CAAC,KAAK;QACjB,MAAM,EAAE,IAAI,CAAC,MAAM;QACnB,QAAQ,EAAE,IAAI,CAAC,QAAQ;QACvB,QAAQ,EAAE,IAAI,CAAC,QAAQ;QACvB,UAAU,EAAE,IAAI,CAAC,UAAU;QAC3B,QAAQ,EAAE,IAAI,CAAC,QAAQ;QACvB,KAAK,EAAE,IAAI,CAAC,KAAK;KAClB,CAAC;AACJ,CAAC;AAWD,MAAM,UAAU,WAAW,CACzB,KAAiB,EACjB,OAA2B;IAE3B,MAAM,aAAa,GAAG,MAAM,CAAC,WAAW,CACtC,KAAK,EACL,OAAO,CAAC,MAAM,EACd,OAAO,CAAC,aAAa,IAAI,KAAK,CAC/B,CAAC;IACF,OAAO,aAAa,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC;AACvC,CAAC;AAED,eAAe,SAAS,CAAC"}
|
package/dist/native.d.ts
CHANGED
|
@@ -20,6 +20,14 @@ export interface LiteParseNativeConfig {
|
|
|
20
20
|
numWorkers?: number;
|
|
21
21
|
ocrFailureFatal?: boolean;
|
|
22
22
|
ocrHedgeDelaysMs?: number[];
|
|
23
|
+
emitWordBoxes?: boolean;
|
|
24
|
+
}
|
|
25
|
+
export interface NativeWordBox {
|
|
26
|
+
text: string;
|
|
27
|
+
x: number;
|
|
28
|
+
y: number;
|
|
29
|
+
width: number;
|
|
30
|
+
height: number;
|
|
23
31
|
}
|
|
24
32
|
export interface NativeTextItem {
|
|
25
33
|
text: string;
|
|
@@ -31,6 +39,7 @@ export interface NativeTextItem {
|
|
|
31
39
|
fontSize?: number;
|
|
32
40
|
confidence?: number;
|
|
33
41
|
rotation?: number;
|
|
42
|
+
words?: NativeWordBox[];
|
|
34
43
|
}
|
|
35
44
|
export interface NativeGraphic {
|
|
36
45
|
kind: string;
|
|
@@ -60,6 +69,7 @@ export interface NativeParsedPage {
|
|
|
60
69
|
width: number;
|
|
61
70
|
height: number;
|
|
62
71
|
text: string;
|
|
72
|
+
markdown: string;
|
|
63
73
|
textItems: NativeTextItem[];
|
|
64
74
|
}
|
|
65
75
|
export interface NativeExtractedImage {
|
package/dist/native.d.ts.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"native.d.ts","sourceRoot":"","sources":["../src/native.ts"],"names":[],"mappings":"AAYA,UAAU,cAAc;IACtB,SAAS,EAAE,KAAK,MAAM,CAAC,EAAE,qBAAqB,KAAK,eAAe,CAAC;IACnE,WAAW,CACT,KAAK,EAAE,cAAc,EAAE,EACvB,MAAM,EAAE,MAAM,EACd,aAAa,CAAC,EAAE,OAAO,GAAG,IAAI,GAC7B,cAAc,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,qBAAqB;IACpC,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,UAAU,CAAC,EAAE,OAAO,CAAC;IACrB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,gBAAgB,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAC1C,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,GAAG,CAAC,EAAE,MAAM,CAAC;IACb,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,YAAY,CAAC,EAAE,OAAO,CAAC;IACvB,qBAAqB,CAAC,EAAE,OAAO,CAAC;IAChC,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,KAAK,CAAC,EAAE,OAAO,CAAC;IAChB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,eAAe,CAAC,EAAE,OAAO,CAAC;IAC1B,gBAAgB,CAAC,EAAE,MAAM,EAAE,CAAC;
|
|
1
|
+
{"version":3,"file":"native.d.ts","sourceRoot":"","sources":["../src/native.ts"],"names":[],"mappings":"AAYA,UAAU,cAAc;IACtB,SAAS,EAAE,KAAK,MAAM,CAAC,EAAE,qBAAqB,KAAK,eAAe,CAAC;IACnE,WAAW,CACT,KAAK,EAAE,cAAc,EAAE,EACvB,MAAM,EAAE,MAAM,EACd,aAAa,CAAC,EAAE,OAAO,GAAG,IAAI,GAC7B,cAAc,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,qBAAqB;IACpC,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,UAAU,CAAC,EAAE,OAAO,CAAC;IACrB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,gBAAgB,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAC1C,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,GAAG,CAAC,EAAE,MAAM,CAAC;IACb,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,YAAY,CAAC,EAAE,OAAO,CAAC;IACvB,qBAAqB,CAAC,EAAE,OAAO,CAAC;IAChC,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,KAAK,CAAC,EAAE,OAAO,CAAC;IAChB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,eAAe,CAAC,EAAE,OAAO,CAAC;IAC1B,gBAAgB,CAAC,EAAE,MAAM,EAAE,CAAC;IAC5B,aAAa,CAAC,EAAE,OAAO,CAAC;CACzB;AAED,MAAM,WAAW,aAAa;IAC5B,IAAI,EAAE,MAAM,CAAC;IACb,CAAC,EAAE,MAAM,CAAC;IACV,CAAC,EAAE,MAAM,CAAC;IACV,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,MAAM,CAAC;CAChB;AAED,MAAM,WAAW,cAAc;IAC7B,IAAI,EAAE,MAAM,CAAC;IACb,CAAC,EAAE,MAAM,CAAC;IACV,CAAC,EAAE,MAAM,CAAC;IACV,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,MAAM,CAAC;IACf,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,KAAK,CAAC,EAAE,aAAa,EAAE,CAAC;CACzB;AAED,MAAM,WAAW,aAAa;IAC5B,IAAI,EAAE,MAAM,CAAC;IACb,EAAE,CAAC,EAAE,MAAM,CAAC;IACZ,EAAE,CAAC,EAAE,MAAM,CAAC;IACZ,EAAE,CAAC,EAAE,MAAM,CAAC;IACZ,EAAE,CAAC,EAAE,MAAM,CAAC;IACZ,CAAC,CAAC,EAAE,MAAM,CAAC;IACX,CAAC,CAAC,EAAE,MAAM,CAAC;IACX,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,SAAS,CAAC,EAAE,OAAO,CAAC;IACpB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED,MAAM,WAAW,eAAe;IAC9B,UAAU,EAAE,MAAM,CAAC;IACnB,SAAS,EAAE,MAAM,CAAC;IAClB,UAAU,EAAE,MAAM,CAAC;IACnB,SAAS,EAAE,cAAc,EAAE,CAAC;IAC5B,QAAQ,CAAC,EAAE,aAAa,EAAE,CAAC;CAC5B;AAED,MAAM,WAAW,gBAAgB;IAC/B,OAAO,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,MAAM,CAAC;IACf,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,EAAE,MAAM,CAAC;IACjB,SAAS,EAAE,cAAc,EAAE,CAAC;CAC7B;AAED,MAAM,WAAW,oBAAoB;IACnC,EAAE,EAAE,MAAM,CAAC;IACX,IAAI,EAAE,MAAM,CAAC;IACb,MAAM,EAAE,MAAM,CAAC;IACf,KAAK,EAAE,MAAM,CAAC;CACf;AAED,MAAM,WAAW,iBAAiB;IAChC,KAAK,EAAE,gBAAgB,EAAE,CAAC;IAC1B,IAAI,EAAE,MAAM,CAAC;IACb,MAAM,EAAE,oBAAoB,EAAE,CAAC;CAChC;AAED,MAAM,WAAW,sBAAsB;IACrC,OAAO,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,MAAM,CAAC;IACf,WAAW,EAAE,MAAM,CAAC;CACrB;AAED,MAAM,WAAW,yBAAyB;IACxC,UAAU,EAAE,MAAM,CAAC;IACnB,UAAU,EAAE,MAAM,CAAC;IACnB,YAAY,EAAE,MAAM,CAAC;IACrB,oBAAoB,EAAE,OAAO,CAAC;IAC9B,eAAe,EAAE,MAAM,CAAC;IACxB,aAAa,EAAE,MAAM,CAAC;IACtB,oBAAoB,EAAE,MAAM,CAAC;IAC7B,aAAa,EAAE,OAAO,CAAC;IACvB,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAC7B,SAAS,EAAE,OAAO,CAAC;IACnB,QAAQ,EAAE,MAAM,CAAC;IACjB,QAAQ,EAAE,OAAO,CAAC;IAClB,OAAO,EAAE,MAAM,EAAE,CAAC;CACnB;AAED,MAAM,WAAW,eAAe;IAC9B,KAAK,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM,GAAG,OAAO,CAAC,iBAAiB,CAAC,CAAC;IAC1D,UAAU,CAAC,KAAK,EAAE,eAAe,EAAE,GAAG,iBAAiB,CAAC;IACxD,SAAS,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM,GAAG,OAAO,CAAC,yBAAyB,EAAE,CAAC,CAAC;IACxE,UAAU,CACR,KAAK,EAAE,MAAM,GAAG,MAAM,EACtB,WAAW,CAAC,EAAE,MAAM,EAAE,GAAG,IAAI,GAC5B,OAAO,CAAC,sBAAsB,EAAE,CAAC,CAAC;IACrC,MAAM,CAAC,MAAM,EAAE,iBAAiB,GAAG,MAAM,CAAC;IAC1C,QAAQ,CAAC,MAAM,EAAE,qBAAqB,CAAC;CACxC;AAkED,eAAO,MAAM,MAAM,gBAAe,CAAC"}
|
package/dist/native.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"native.js","sourceRoot":"","sources":["../src/native.ts"],"names":[],"mappings":"AAAA,2FAA2F;AAC3F,EAAE;AACF,kFAAkF;AAClF,gGAAgG;AAEhG,OAAO,EAAE,aAAa,EAAE,MAAM,aAAa,CAAC;AAC5C,OAAO,EAAE,IAAI,EAAE,OAAO,EAAE,MAAM,WAAW,CAAC;AAC1C,OAAO,EAAE,aAAa,EAAE,MAAM,UAAU,CAAC;AAEzC,MAAM,OAAO,GAAG,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;AAC/C,MAAM,SAAS,GAAG,OAAO,CAAC,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;
|
|
1
|
+
{"version":3,"file":"native.js","sourceRoot":"","sources":["../src/native.ts"],"names":[],"mappings":"AAAA,2FAA2F;AAC3F,EAAE;AACF,kFAAkF;AAClF,gGAAgG;AAEhG,OAAO,EAAE,aAAa,EAAE,MAAM,aAAa,CAAC;AAC5C,OAAO,EAAE,IAAI,EAAE,OAAO,EAAE,MAAM,WAAW,CAAC;AAC1C,OAAO,EAAE,aAAa,EAAE,MAAM,UAAU,CAAC;AAEzC,MAAM,OAAO,GAAG,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;AAC/C,MAAM,SAAS,GAAG,OAAO,CAAC,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;AAuI1D,SAAS,UAAU;IACjB,uDAAuD;IACvD,MAAM,OAAO,GAA2B;QACtC,YAAY,EAAE,kCAAkC;QAChD,cAAc,EAAE,oCAAoC;QACpD,eAAe,EAAE,qCAAqC;QACtD,gBAAgB,EAAE,sCAAsC;QACxD,iBAAiB,EAAE,uCAAuC;QAC1D,kBAAkB,EAAE,wCAAwC;QAC5D,gBAAgB,EAAE,sCAAsC;QACxD,kBAAkB,EAAE,wCAAwC;KAC7D,CAAC;IAEF,sCAAsC;IACtC,MAAM,QAAQ,GAAG,OAAO,CAAC,QAAQ,CAAC;IAClC,MAAM,IAAI,GAAG,OAAO,CAAC,IAAI,CAAC;IAE1B,MAAM,UAAU,GAAa,EAAE,CAAC;IAChC,IAAI,QAAQ,KAAK,OAAO,EAAE,CAAC;QACzB,2BAA2B;QAC3B,UAAU,CAAC,IAAI,CAAC,GAAG,QAAQ,IAAI,IAAI,MAAM,CAAC,CAAC;QAC3C,UAAU,CAAC,IAAI,CAAC,GAAG,QAAQ,IAAI,IAAI,OAAO,CAAC,CAAC;IAC9C,CAAC;SAAM,IAAI,QAAQ,KAAK,OAAO,EAAE,CAAC;QAChC,UAAU,CAAC,IAAI,CAAC,GAAG,QAAQ,IAAI,IAAI,OAAO,CAAC,CAAC;IAC9C,CAAC;SAAM,CAAC;QACN,UAAU,CAAC,IAAI,CAAC,GAAG,QAAQ,IAAI,IAAI,EAAE,CAAC,CAAC;IACzC,CAAC;IAED,KAAK,MAAM,GAAG,IAAI,UAAU,EAAE,CAAC;QAC7B,MAAM,GAAG,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC;QACzB,IAAI,GAAG,EAAE,CAAC;YACR,IAAI,CAAC;gBACH,OAAO,OAAO,CAAC,GAAG,CAAC,CAAC;YACtB,CAAC;YAAC,MAAM,CAAC;gBACP,0BAA0B;YAC5B,CAAC;QACH,CAAC;IACH,CAAC;IAED,kDAAkD;IAClD,8DAA8D;IAC9D,MAAM,UAAU,GAAG,CAAC,SAAS,EAAE,IAAI,CAAC,SAAS,EAAE,IAAI,CAAC,EAAE,IAAI,CAAC,SAAS,EAAE,IAAI,EAAE,IAAI,CAAC,CAAC,CAAC;IACnF,4EAA4E;IAC5E,MAAM,SAAS,GAAG;QAChB,GAAG,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,aAAa,CAAC,OAAO,CAAC;QAC/C,aAAa,QAAQ,IAAI,IAAI,OAAO;QACpC,gBAAgB;KACjB,CAAC;IACF,KAAK,MAAM,GAAG,IAAI,UAAU,EAAE,CAAC;QAC7B,KAAK,MAAM,QAAQ,IAAI,SAAS,EAAE,CAAC;YACjC,IAAI,CAAC;gBACH,OAAO,OAAO,CAAC,IAAI,CAAC,GAAG,EAAE,QAAQ,CAAC,CAAC,CAAC;YACtC,CAAC;YAAC,MAAM,CAAC;gBACP,WAAW;YACb,CAAC;QACH,CAAC;IACH,CAAC;IAED,MAAM,IAAI,KAAK,CACb,oCAAoC,QAAQ,IAAI,IAAI,IAAI;QACtD,sDAAsD,CACzD,CAAC;AACJ,CAAC;AAED,MAAM,CAAC,MAAM,MAAM,GAAG,UAAU,EAAE,CAAC"}
|
|
Binary file
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@llamaindex/liteparse",
|
|
3
|
-
"version": "2.
|
|
3
|
+
"version": "2.4.1",
|
|
4
4
|
"description": "Fast, lightweight PDF and document parsing with spatial text extraction",
|
|
5
5
|
"type": "module",
|
|
6
6
|
"main": "./dist/lib.js",
|
|
@@ -54,13 +54,13 @@
|
|
|
54
54
|
"typescript": "~5.9.2"
|
|
55
55
|
},
|
|
56
56
|
"optionalDependencies": {
|
|
57
|
-
"@llamaindex/liteparse-darwin-arm64": "2.
|
|
58
|
-
"@llamaindex/liteparse-darwin-x64": "2.
|
|
59
|
-
"@llamaindex/liteparse-linux-arm64-gnu": "2.
|
|
60
|
-
"@llamaindex/liteparse-linux-x64-gnu": "2.
|
|
61
|
-
"@llamaindex/liteparse-win32-arm64-msvc": "2.
|
|
62
|
-
"@llamaindex/liteparse-win32-x64-msvc": "2.
|
|
63
|
-
"@llamaindex/liteparse-linux-x64-musl": "2.
|
|
57
|
+
"@llamaindex/liteparse-darwin-arm64": "2.4.1",
|
|
58
|
+
"@llamaindex/liteparse-darwin-x64": "2.4.1",
|
|
59
|
+
"@llamaindex/liteparse-linux-arm64-gnu": "2.4.1",
|
|
60
|
+
"@llamaindex/liteparse-linux-x64-gnu": "2.4.1",
|
|
61
|
+
"@llamaindex/liteparse-win32-arm64-msvc": "2.4.1",
|
|
62
|
+
"@llamaindex/liteparse-win32-x64-msvc": "2.4.1",
|
|
63
|
+
"@llamaindex/liteparse-linux-x64-musl": "2.4.1"
|
|
64
64
|
},
|
|
65
65
|
"engines": {
|
|
66
66
|
"node": ">=18.0.0"
|